In our original proposal, we sought to define and characterize modified citrus pectin, an inhibitor of metastasis. The results of these studies have defined a subcomponent of MCP which inhibits human prostate cancer metastasis in athymic animals and can be made in quantities sufficient for human clinical trials. An Investigational New Drug application will be submitted by March 1, 1998. The ultimate goal of this proposal is to develop a therapy to inhibit the process of bone metastasis by human prostate cancer cells. In virtually 100% of patients with advanced disease, prostate cancer metastasizes to osseous sites and causes osteoblastic growth. The mechanisms which contribute to bone metastasis are poorly understood, however, it has been hypothesized that the in situ environment of the bone provides a favorable growth advantage for prostate cancer cells which non-selectively seed the bone marrow from the bloodstream. An alternative hypothesis suggests that prostate cancer cells preferentially bind to bone marrow endothelial cells as compared to endothelial cells lining the blood vessels of other organs. The development of a workable model of bone endothelial is a logical step in the investigation of prostate cancer bone metastasis and in our preliminary results we report the isolation and characterization of a human bone marrow endothelial cell line (HBME-1). In an vitro assays, prostate cancer cells adhered preferentially to HBME-1 cells are compared to endothelium derived from other organs. These data suggest that the propensity of prostate cancer cells to establish themselves in bone is due, at least in part, to their selective adhesion to human bone marrow endothelial cells. We further demonstrate that this preferential binding of prostate cancer cells is mediated, at least in part, by the presence of galectin-3, a galactose binding lectin, on the surface of the human prostate cancer cells. While this adhesion can be inhibited by complex carbohydrates such as modified citrus pectin and antibodies to galectin-3, the ligand on HBME-1 cells to which galectin-3 binds remains unknown. Since galectin-3 is up-regulated in multiple metastatic tumor types, not all of which metastasize to bone, we hypothesize that the identification of the HBME-1 receptor is the key to understanding the process of prostate cancer metastasis to bone and to the development of bone specific anti- adhesion strategies. Over the next two years, we proposed to: 1) Identify the galectin-3 ligand on human bone marrow endothelial cells. 2) Characterize this HBME-1/galectin-3 ligand to allow the development of anti-adhesion inhibitors.